Reward shaping-based deep reinforcement learning for look-ahead dispatch with rolling-horizon

The increasing penetration of renewable energy exacerbates the challenges in designing an effective and adaptable model-driven Look-ahead Dispatch (LAD) method. Recently, deep reinforcement learning (DRL) methods show enormous potential in developing a dispatching agent with self-learning ability, a...

Full description

Saved in:
Bibliographic Details
Main Authors: Hongsheng Xu, Yungui Xu, Ke Wang, Yaping Li, Abdullah Al Ahad
Format: Article
Language:English
Published: Elsevier 2025-07-01
Series:International Journal of Electrical Power & Energy Systems
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S0142061525002248
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849321637560713216
author Hongsheng Xu
Yungui Xu
Ke Wang
Yaping Li
Abdullah Al Ahad
author_facet Hongsheng Xu
Yungui Xu
Ke Wang
Yaping Li
Abdullah Al Ahad
author_sort Hongsheng Xu
collection DOAJ
description The increasing penetration of renewable energy exacerbates the challenges in designing an effective and adaptable model-driven Look-ahead Dispatch (LAD) method. Recently, deep reinforcement learning (DRL) methods show enormous potential in developing a dispatching agent with self-learning ability, attributed to their superior generalization, adaptability, and computational efficiency. However, existing DRL-based LAD methods overlook the discounting effect when calculating the immediate total reward for LAD and lack attention to trial-and-error reward design and expected discounted returns that could reflect the true performance metrics of LAD. Therefore, this paper proposes novel reward shaping (RS)-based DRL algorithms for the rolling-horizon LAD problem. We propose the method for accurately estimating the look-ahead discounted factor that best matches different look-ahead horizons (LAHs). The shaped reward functions are designed and an RS-based regularization is also proposed by employing a potential function. Case studies on the SG 126-bus and IEEE 118-bus systems demonstrate the effectiveness of the proposed improved measures, as well as the superiority and adaptability of the proposed improved DRL algorithms in training and testing performance.© 2017 Elsevier Inc. All rights reserved.
format Article
id doaj-art-c4f953377bcc4a6bb2b786d303541b1a
institution Kabale University
issn 0142-0615
language English
publishDate 2025-07-01
publisher Elsevier
record_format Article
series International Journal of Electrical Power & Energy Systems
spelling doaj-art-c4f953377bcc4a6bb2b786d303541b1a2025-08-20T03:49:41ZengElsevierInternational Journal of Electrical Power & Energy Systems0142-06152025-07-0116811067310.1016/j.ijepes.2025.110673Reward shaping-based deep reinforcement learning for look-ahead dispatch with rolling-horizonHongsheng Xu0Yungui Xu1Ke Wang2Yaping Li3Abdullah Al Ahad4School of Electrical and Power Engineering, Hohai University, Nanjing 211100, ChinaSchool of Electrical and Power Engineering, Hohai University, Nanjing 211100, ChinaSchool of Electrical and Power Engineering, Hohai University, Nanjing 211100, China; Corresponding author.Department of Power Automation, China Electric Power Research Institute, Nanjing 210000, ChinaSchool of Electrical and Power Engineering, Hohai University, Nanjing 211100, ChinaThe increasing penetration of renewable energy exacerbates the challenges in designing an effective and adaptable model-driven Look-ahead Dispatch (LAD) method. Recently, deep reinforcement learning (DRL) methods show enormous potential in developing a dispatching agent with self-learning ability, attributed to their superior generalization, adaptability, and computational efficiency. However, existing DRL-based LAD methods overlook the discounting effect when calculating the immediate total reward for LAD and lack attention to trial-and-error reward design and expected discounted returns that could reflect the true performance metrics of LAD. Therefore, this paper proposes novel reward shaping (RS)-based DRL algorithms for the rolling-horizon LAD problem. We propose the method for accurately estimating the look-ahead discounted factor that best matches different look-ahead horizons (LAHs). The shaped reward functions are designed and an RS-based regularization is also proposed by employing a potential function. Case studies on the SG 126-bus and IEEE 118-bus systems demonstrate the effectiveness of the proposed improved measures, as well as the superiority and adaptability of the proposed improved DRL algorithms in training and testing performance.© 2017 Elsevier Inc. All rights reserved.http://www.sciencedirect.com/science/article/pii/S0142061525002248Look-ahead dispatchRolling-horizonDeep reinforcement learningReward shapingSoft actor-critic
spellingShingle Hongsheng Xu
Yungui Xu
Ke Wang
Yaping Li
Abdullah Al Ahad
Reward shaping-based deep reinforcement learning for look-ahead dispatch with rolling-horizon
International Journal of Electrical Power & Energy Systems
Look-ahead dispatch
Rolling-horizon
Deep reinforcement learning
Reward shaping
Soft actor-critic
title Reward shaping-based deep reinforcement learning for look-ahead dispatch with rolling-horizon
title_full Reward shaping-based deep reinforcement learning for look-ahead dispatch with rolling-horizon
title_fullStr Reward shaping-based deep reinforcement learning for look-ahead dispatch with rolling-horizon
title_full_unstemmed Reward shaping-based deep reinforcement learning for look-ahead dispatch with rolling-horizon
title_short Reward shaping-based deep reinforcement learning for look-ahead dispatch with rolling-horizon
title_sort reward shaping based deep reinforcement learning for look ahead dispatch with rolling horizon
topic Look-ahead dispatch
Rolling-horizon
Deep reinforcement learning
Reward shaping
Soft actor-critic
url http://www.sciencedirect.com/science/article/pii/S0142061525002248
work_keys_str_mv AT hongshengxu rewardshapingbaseddeepreinforcementlearningforlookaheaddispatchwithrollinghorizon
AT yunguixu rewardshapingbaseddeepreinforcementlearningforlookaheaddispatchwithrollinghorizon
AT kewang rewardshapingbaseddeepreinforcementlearningforlookaheaddispatchwithrollinghorizon
AT yapingli rewardshapingbaseddeepreinforcementlearningforlookaheaddispatchwithrollinghorizon
AT abdullahalahad rewardshapingbaseddeepreinforcementlearningforlookaheaddispatchwithrollinghorizon